CONTACT PHENOMENA IN TELEPHONE SWITCHING CIRCUITS 59 



toothed oscillation of from 10 to 50 volts peak-to-peak. Part of an 

 "A" type transient showing this peculiarity is illustrated by the 

 oscillogram of Fig. 25. This appears to be a relaxation oscillation 

 such as is commonly produced in ionized gas tubes, riding on the normal 

 300-volt axis of the gas glow discharge. The conditions which lead 

 to the occurrence of this oscillation at atmospheric pressure have not 

 been identified, but it is found to be quite stable in some cases where 

 contacts have been sealed in a mixture of air and gases at about half 

 atmospheric pressure. 



A typical "A" transient is shown in detail in Figs. 26, 27, and 28. 

 The circuit consisted of a 250-ohm relay connected to the contacts by 

 10 ft. of wire, the battery being 50 volts as usual. Figure 26 shows 

 the voltage of the early part of the transient during which rapid 

 sparkovers are interspersed with two brief periods during which a gas 

 glow discharge was established but not maintained. Figure 27 shows 

 the final sparkover before the establishment of the glow discharge at 

 about 300 volts. A group of the current pulses corresponding to the 

 initial part of the sparkover stage is shown in Fig. 28. These are 

 complicated by the line oscillations (which should be of about 30 

 megacycles frequency) and appear to last less than 0.1 microsecond. 



It may be seen that the individual sparkovers at the start of the 

 "A" transient are somewhat different in form from those of the "B" 

 transient. The voltage reaches 320 volts in a microsecond or so, and 

 in some cases collapses to zero or beyond immediately. There are 

 sometimes indications of arcing periods lasting much less than 0.1 

 microsecond and the voltage recovers with oscillations of the line wire 

 but the duration of the phenomenon is too brief for very accurate 

 analysis. But in many cases, the voltage, having reached its peak, 

 drops to an intermediate value of 280 volts and recovers to 320 volts 

 before it collapses. This is probably due to the temporary formation 

 of the nitrogen glow discharge, which is finally established and main- 

 tained during the remainder of the contact opening when for some 

 reason the sparkover does not occur. In cases where the contacts are 

 on the verge of producing a "B" transient the voltage may rise to 

 500 volts and then collapse to the 300 volts of the gas glow discharge. 



It is very interesting to set up a circuit which will cause the "A" 

 transient to predominate, and start operating freshly filed contacts 

 several times a second observing the transient voltage at contact 

 opening on the oscilloscope. The first transient will always be of the 

 "B" type. Usually the first few dozen will also. However, after a 

 while one of the transients will show a flat top at about 300 volts for 

 a very brief period and this tendency increases until finally a complete 



